An artificial olfactory system is a device that is capable of detecting a wide variety of analytes in fluids such as vapors, gases and liquids. The device comprises an array of sensors that in the presence of an analyte produces a response. The device produces a unique signature output for a particular analyte. Using pattern recognition algorithms, the output signature, such as an electrical response, can be correlated and compared to a particular analyte or mixture of substances that are known. By comparing the unknown signature with the stored or known signatures, the analyte can be detected, identified and quantified.
There are many instances where it is desirable to measure trace amounts of analytes. However, in certain instances, the analytes are found at levels that are too low to register a robust signal by direct exposure to currently available sensors. In headspace analysis of applications in agricultural, wine, tobacco, perfume, plastics, and the food industries, the detection and classification of trace levels of gases are present in the sub part per million (ppm) range, making detection difficult. Moreover, in residue analysis of pesticides on crops, the trace levels of certain herbicides must meet federal guidelines. For certain crops, these residues are present on the crops in the part per billion levels (ppb).
Another potential application wherein the detection of trace levels of analytes is important is the diagnosis of patients' conditions from an analysis of their breath. Marker gases such as hydrogen sulfide and methyl mercaptan, which are important in diagnosing the presence of oral or lung conditions from the breath of human patients, often exist in concentrations of 0.01-1 parts per million (or lower). However, the threshold detection levels of currently known sensors are in the range of 1-100 parts per million.
Currently, the most widely used device for detecting oral malodors is the Halimeter, which is commercially available from Interscan Corp. (Chatsworth, Calif.). Using an electrochemical cell that is sensitive to volatile sulfur compounds (VSC), the device can oxidize the VSC at the anode according to the following reactions: EQU H.sub.2 S.fwdarw.S+2H.sup.+ +2e.sup.- EQU 2CH.sub.3 --SH.fwdarw.CH.sub.3 --S--S--CH.sub.3 +S+2H.sup.+ +2e.sup.- EQU CH.sub.3 --S--CH.sub.3 +2H.sub.2 O.fwdarw.2CH.sub.3 OH+S+2H.sup.+ +2e.sup.-
However, one obvious drawback is that the Halimeter cannot distinguish between volatile sulfur compounds. Similarly, other volatile substances can interfere with the readings of the VSC.
A second device for the detection of breath and odors associated therewith is based on zinc-oxide thin film semiconductor technology and has recently been developed for measuring VSC (see, Shimura M et al., J Periodontol. 67:396-402 (1994)). New Cosmos Electric Co. (Osaka, Japan) manufactures this device. This device, however, is limited because it is susceptible to interference from organic vapors unrelated to oral malodor (see, Yaegaki K, In Rosenberg, "Bad Breath: Research Perspectives," Proceedings of the First International Workshop on Oral Malodor, Ramot Publishing, Tel Aviv University pp. 41-54 (1993)) 87-108 (1995)).
Another analysis for breath detection is a test based on the enzyme substrate benzoyl-DL-arginine-naphthylamine (BANA) (see, Loesche et al., J. Clin. Microbiol. 28:1551-1559 (1991); and Loesche et al. J Periodontol., 61:189-196 (1991)). This test is marketed under the brand name Peroscan and is available from Oral-B Laboratories (Redwood City, Calif.). Scrapings from the tongue, saliva, or plaque samples are deposited directly on a reagent card. Following substrate addition, a blue spot develops if anaerobes are present. Studies have shown that BANA results are not highly correlated with VSC measurements and that the test is often detecting other analytes (see, Kozlovsky et al., J Dent. Res., 73: 103-1042 (1994)).
It has been estimated that at least 50% of the population suffers from chronic oral malodor (see, Bosy, J. Can. Dent. Assoc. 63:196-201 (1997)). A significant fraction of the population is worried about bad breath, even though there is usually no underlying disease (see, Iwakura et al, J. Dent. Res. 7:1568-1574 1 (1994)). Food, of course, is another cause of oral malodors. However, there are many people who have an unwarranted phobia of bad breath. The size of the market for breath fresheners, chewing gums, and mouth rinses is an indicator of this propensity.
In addition to mammalian breath measurements, respiratory devices for anesthetic and respiratory gas mixtures must be monitored at very low concentrations of analytes. Medical devices mix the anesthetic with breathing gas prior to delivery to the patient. In an anesthetic device, it is imperative that the concentration of the anesthetic, gas flow and amounts of the mixture and starting gases be known with certainty. In most instances, the anesthetic amounts are at very low concentration levels.
One approach to increase sensitivity to certain analytes is to use selective filters or membranes. For instance, U.S. Pat. No. 5,841,021, which issued to De Castro et al., discloses an electrochemical gas sensor that has a catalytically active sensor electrode, a reference electrode and a permselective filter or membrane layer. The filter is made of a material that provides for molecular specificity of certain gases, such as carbon monoxide. The membrane allows the sensors to be selective to the chemical analyte of interest. The filter only allows the analytes of interest to contact the sensor. By removing interfering substances through filtration, the sensor becomes more selective and thus sensitive to the analyte of interest.
In addition, U.S. Pat. No. 5,057,436, which issued to Ball, discloses a method and apparatus for detection of toxic gases, such as ammonia, using a metal oxide semiconductor and an electrochemical sensor. Disposed between the two sensors is an absorber having an absorbent that reacts with ammonia.
In view of the foregoing, what is needed in the art is a vapor concentrator for an array of sensors, especially for an electronic nose sensor array. In addition, methods are needed to detect odors and diagnose medical conditions. The present invention fulfills these and other needs.